Is Timing of Cranioplasty Following Posttraumatic Craniectomy Related to Neurological Outcome?

ORIGINAL RESEARCH

International Journal of Surgery 11 (2013) 886e890

Contents lists available at ScienceDirect

International Journal of Surgery

journal homepage: www.journal-surgery.net

Original research Is timing of cranioplasty following posttraumatic craniectomy related to neurological outcome?

Yu-Hua Huang a,b,*, Tao-Chen Lee a, Ka-Yen Yang a, Chen-Chieh Liao a a Department of Neurosurgery, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Kaohsiung, Taiwan b Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan article info abstract

Article history: Background: With the use of decompressive craniectomy for traumatic brain injury (TBI) come a cor- Received 27 April 2013 responding number of cranioplasties. TBI causes dynamic processes to commence or change during the Received in revised form period from injury to recovery; hence, the role of the timing of surgical intervention should be 18 June 2013 emphasized. Accepted 27 July 2013 Aims: We attempt to identify the relationship between the timing of cranioplasty and neurological Available online 7 August 2013 outcomes following posttraumatic craniectomy. Methods: In this 3-year retrospective study, 105 patients undergoing decompressive craniectomies and Keywords: ’ Cranioplasty subsequent cranioplasties for TBI were enrolled. We documented the patients demographic information, Decompressive craniectomy including Glasgow Coma Scale (GCS) at admission for trauma. The follow-up period was terminated by Timing death or a minimum of 6 months after TBI. Glasgow Outcome Scale (GOS) at the end of follow-up was Outcomes used as an outcome measure. Unfavorable outcome was defined as a GOS score of 1e3. Complications Results: The 105 patients included 71 male and 34 female subjects. The mean age was 41.94 19.73 years. Neurological assessment showed that admission GCS was 8.50 3.15, on average. The mean time interval between cranioplasty and craniectomy was 78.84 49.04 days (range, 13e245 days). Univariate logistic regression analysis showed that the association between the timing of cranioplasty and unfavorable outcomes was not statistically significant (odds ratio ¼ 1.005, confidence interval 0.997e1.013; p ¼ 0.195). Conclusion: The timing of cranioplasty following posttraumatic craniectomy was not related to the neurological outcomes of TBI. Despite the limitations of the retrospective design, the analyses provide preliminary information to elucidate the question. Ó 2013 Surgical Associates Ltd. Published by Elsevier Ltd. All rights reserved.

e 1. Introduction and hydrocephalus.3 5 In fact, the neurological outcomes are of fundamental importance to the heterogeneous condition of TBI. In patients with traumatic brain injury (TBI), decompressive Age, Glasgow Coma Scale (GCS), pupil reactivity, computed to- craniectomy (DC) is an effective means of controlling high mography (CT) abnormalities, and systemic insults (hypoxia and intracranial pressure.1,2 Though DC is a life-saving measure, it hypotension) are established relevant factors related to outcomes may necessitate a second cranioplasty surgery for survivors to in patients with moderate and severe TBI.6 Thus far, the timing of replace the bone. Repeat surgery inherently increases the risks cranioplasty in relation to neurological outcomes in the popu- of morbidity or mortality, and furthermore, the appropriate lation undergoing craniectomies for TBI has not been well timing of this staged procedure remains a problem. When analyzed. The issue of the optimal time interval is particularly considering the timing of cranioplasty, predominant concerns in meaningful not only for patients and their families, but also for the literature include residual brain swelling, risk of infection, the clinicians who have to make a decision on timing. Consequently, in this study, we retrospectively collected the clinical details of patients undergoing DC for TBI, and analyzed * Corresponding author. Department of Neurosurgery, Kaohsiung Chang Gung neurological outcomes after cranioplasty procedures. We attemp- Memorial Hospital, 123, Ta Pei Road, Niao Sung District, Kaohsiung, Taiwan. Tel.: þ886 7 7317123x8011; fax: þ886 7 7354309. ted to identify the relationship between the timing of cranioplasty E-mail address: [email protected] (Y.-H. Huang). and prognosis following posttraumatic craniectomy.

Y.-H. Huang et al. / International Journal of Surgery 11 (2013) 886e890 887

2. Materials and methods 2.7. Statistical analysis

2.1. Patient collection Data were analyzed using SPSS version 12.0 (SPSS Inc., Chicago, IL). Categorical and continuous variables are presented in percentage and mean, respectively. Lo- From January 2006 to December 2008, 201 patients underwent DC for TBI at gistic regression was performed to identify factors associated with unfavorable Kaohsiung Chang Gung Memorial Hospital, a 2754-bed acute care medical center in neurological outcomes. Results are expressed as odds ratios with 95% confidence southern Taiwan. We excluded 59 patients who died before cranioplasty, 24 who intervals. A p value of less than 0.05 was considered statistically significant. refused a second operation because of risks, and 13 that were lost to follow-up. Thus, a total of 105 patients undergoing subsequent cranial reconstruction were enrolled for the analysis. The patients’ charts were retrospectively reviewed after approval by 3. Results our Institutional Review Board. We documented the patients’ demographic data, Injury Severity Score (ISS), pupil reactivity, GCS at admission, and the time interval General epidemiologic data and stratification by neurological between cranioplasty and DC. outcomes are listed in Table 1. The 105 patients who underwent DC and subsequent cranioplasty included 71 male and 34 female 2.2. Neuroimaging evaluation subjects. The mean age was 41.94 19.73 years. The mechanisms of head injury were as follows: 86 traffic accidents, 18 fall accidents, The patients underwent CT of the brain soon after arrival at the emergency and 1 other cause. At admission, the mean ISS was 25.76 6.63 room. Follow-up CT scans were obtained in cases with acute onset of focal neuro- e logical deficits, progressively disturbed consciousness, or absence of neurological (range, 16 57). Neurological assessment showed that the mean improvement. The following features on CT scans before DC were evaluated: (a) the GCS was 8.50 3.15. Pupil examination identified 68 patients with presence of subdural hemorrhage, subarachnoid hemorrhage, intraventricular 2 reacting pupils, and 37 patients with one or both non-reacting hemorrhage, epidural hemorrhage, contusional hemorrhage, or skull fracture; (b) pupils. the status of basal cisterns, subdivided into normal, compressed, or absent; (c) the The individual CT scan features included 86.7% SDH, 81.9% SAH, degree of midline shift measured as the deviation of the septum pellucidum from the central position. 6.7% IVH, 60.0% contusional hemorrhage, 16.2% EDH, and 61.9% skull fracture. The status of the basal cistern was normal, compressed, and absent in 17 (16.2%), 68 (64.8%), and 20 (19.0%) pa- 2.3. DC indications and techniques tients, respectively. The mean midline shift was 7.49 5.09 mm The clinical indications and roles of DC for TBI were as follows.7 Primary DC was (range, 0e22.20 mm). defined as surgical decompression, with or without brain tissue removal, primarily Primary DC was performed in 102 of the 105 patients, and for the evacuation of any type of intradural lesion. The aim of prophylactic cra- secondary DC for the control of refractory intracranial hypertension niectomy was to avoid expected postsurgical increases in intracranial pressure. Secondary DC was performed in patients whose high intracranial pressure was refractory to medical treatment. Any patient who had undergone an initial surgical Table 1 procedure to evacuate a space-occupying lesion, and subsequently developed Clinical characteristics of 105 patients in terms of neurological outcome. delayed brain swelling, was also indicated for secondary DC. When brain swelling Total cases Unfavorable Favorable was limited to one cerebral hemisphere, unilateral hemicraniectomy was per- N ¼ 105 outcome outcome formed. For bilateral hemispheres or frontal swelling, bilateral hemicraniectomy or N ¼ 37 N ¼ 68 bifrontal craniectomy was chosen, respectively. The dura mater was opened, and the opening was extended to the bone margins. The brain surface was covered loosely Mean age (year) 41.94 19.73 51.89 18.52 36.53 18.31 by the remaining dura or artificial dural substitutes. The bone flaps were stored in a Gender deep freezer at 75 C for various time periods. Male 71 (67.6%) 25 (67.6%) 46 (67.6%) Female 34 (32.4%) 12 (32.4%) 22 (32.4%) Mechanism of head injury 2.4. Cranioplasty indications and techniques Traffic accident 86 (81.9%) 27 (73.0%) 59 (86.8%) Fall accident 18 (17.1%) 10 (27.0%) 8 (11.8%) All subsequent cranioplasties were performed with autologous bone flaps if the Others 1 (1.0%) 0 (0.0%) 1 (1.5%) presence of slack brain and medical status allowed reconstruction. The layer for the Mean ISS 25.76 6.63 26.70 7.81 25.25 5.89 replacement of the bone fragment was dissected between the myocutaneous flap Mean GCS at admission 8.50 3.15 7.54 3.48 9.03 2.85 and the dura-like layer (neo-dura) or artificial dural substitute. The margins of the Pupil reactivity at admission craniectomy defects were exposed, and the bone flaps were fixed in their original Both reacting 68 (64.8%) 20 (54.1%) 48 (70.6%) positions using wires or titanium plates with screws. As prophylaxis, all patients One or both non-reacting 37 (35.2%) 17 (45.9%) 20 (29.4%) were intravenously administered with pre- and postoperative antibiotics. Features of CT at admission Subdural hemorrhage 91 (86.7%) 33 (89.2%) 58 (85.3%) Subarachnoid hemorrhage 86 (81.9%) 34 (91.9%) 52 (76.5%) 2.5. Complications related to the timing of cranioplasty Intraventricular hemorrhage 7 (6.7%) 4 (10.8%) 3 (4.4%) Contusional hemorrhage 63 (60.0%) 24 (64.9%) 39 (57.4%) fi Neurosurgical site infections after cranioplasty were classi ed as follows: scalp Epidural hemorrhage 17 (16.2%) 3 (8.1%) 14 (20.6%) fl 8,9 infection; bone ap osteitis; meningitis-ventriculitis; brain abscess or empyema. Skull fracture 65 (61.9%) 17 (45.9%) 48 (70.6%) For each patient, only the most severe infection was recorded. Hydrocephalus before Status of basal cisterns cranioplasty was diagnosed in patients who had a progressive increase in ventricular Normal 17 (16.2%) 2 (5.4%) 15 (22.1%) fl size and needed permanent shunting of the cerebrospinal uid. Correlation of hy- Compressed 68 (64.8%) 26 (70.3%) 42 (61.8%) fi drocephalus with the clinical examination was attempted, but often proved dif cult Absent 20 (19.0%) 9 (24.3%) 11 (16.2%) fi 4 due to the severe neurologic de cit. Midline shift (mm) 7.49 5.09 8.93 6.20 6.70 4.22 Interval from head injury to DC 2.6. Outcome assessment &24 h 84 (80.0%) 29 (78.4%) 55 (80.9%) >24 h 21(20.0%) 8 (21.6%) 13 (19.1%) The follow-up period was terminated by death or a minimum of 6 months after Indications of DC TBI (range: 3e61 months; mean: 25.96 15.61 months). Neurological outcome at Primary 102 (97.1%) 35 (94.6%) 67 (98.5%) the end of follow-up was assessed using the Glasgow Outcome Scale (GOS) as fol- Secondary 3 (2.9%) 2 (5.4%) 1 (1.5%) lows: 1 ¼ death; 2 ¼ persistent vegetative state with inability to interact with the Interval between 78.84 49.04 87.32 55.43 74.22 44.95 environment; 3 ¼ severe disability with inability to live independently, but the cranioplasty and DC (day) ability to follow commands; 4 ¼ moderate disability with the ability to live inde- Neurosurgical site 10 (9.5%) 2 (5.4%) 8 (11.8%) pendently but inability to return to work or school; and 5 ¼ mild or no disability infection after cranioplasty with the ability to return to work or school. For use as dichotomous variables, un- Hydrocephalus 21 (20.0%) 16 (43.2%) 5 (7.4%) favorable and favorable outcomes were defined by GOS scores of 1e3 and 4e5, before cranioplasty respectively. ORIGINAL RESEARCH

888 Y.-H. Huang et al. / International Journal of Surgery 11 (2013) 886e890 was performed in 3 patients. Emergency surgery within 24 h after head trauma was performed in 84 patients, whereas 21 patients were treated with delayed surgery (>24 h after injury). Unilateral frontotemporoparietal hemicraniectomy was performed for 98 patients. Two patients underwent bilateral hemicraniectomy and 5 underwent bifrontal craniectomy. The mean time interval between cranioplasty and DC was 78.84 49.04 days (range, 13e245 days). Fig. 1 shows the distribution of the timing of cranioplasty following DC for the 105 patients. Neurosurgical site infection occurred in 10 patients after cranioplasty and consisted of 4 scalp infections and 6 brain abscess or empyema. The rate of neurosurgical site infection in the unfavorable outcome groups was 5.4%. Hydrocephalus before cranioplasty occurred in 21 patients and all underwent cerebrospinal fluid shunting procedures. The incidence of hydrocephalus in the unfavorable outcome group was 43.2%. The incidence of complications and the time to cranioplasty are shown in Fig. 2. One of the 105 patients died from pneumonia with respiratory failure during hospitalization for cranioplasty, and 6 patients died after discharge. The overall mortality rate was 6.7% at the end of the follow-up period. In addition, 15 patients (14.3%) remained in a vegetative state, and 15 (14.3%) showed severe deficits. Thirty-eight patients (36.2%) had moderate deficits, and 30 (28.6%) showed good recovery and social reintegration. As a Fig. 2. Percentage of complications by stratification of timing of cranioplasty. result, thirty-seven patients (35.2%) had an unfavorable outcome, and 68 (64.8%) had a favorable outcome at the time of the final evaluation. 4. Discussion Logistic regression analysis for unfavorable outcome was performed using the timing of cranioplasty and established predictors DC reduces medically refractory intracranial hypertension, and 12e14 of outcome after TBI (age, GCS, pupil reactivity, and CT abnormal- it is a valuable tool in the management of severe head injury. ity).6,10,11 The univariate model revealed that the timing of cranio- However, the prognosis of patients after DC is highly variable and plasty was not significantly related to unfavorable outcome (odds unsatisfactory for some survivors. In a prospective study, Jiang ratio ¼ 1.005, confidence interval 0.997e1.013; p ¼ 0.195). The et al. enrolled 241 patients undergoing DC for severe TBI, and overall results of univariate logistic regression are presented in reported that at the 6-month follow-up, 71 (30%) had severe 13 Table 2. disability and 9 (4%) were in a vegetative state. In a series of 176 patients, Ho et al. concluded delayed neurological recovery after DC was common, but there were still 34 severely disabled and 5 vegetative patients at the 18-month follow-up.15 Although the factors relevant to outcomes have been documented, all were based mainly on admission characteristics, including age, GCS, and pupil status.16,17 Information obtained during the subsequent clinical course may further contribute to outcome determination. With the use of DC comes a corresponding number of cranioplasties performed to replace the bone defects created. It is un- known whether the variables of this second operation have an impact on neurological outcomes after TBI.

Table 2 Univariate logistic regression analysis of potential predictors and timing of cranioplasty for unfavorable outcome.

Unfavorable outcome P value Odds ratio (95% CI)

Age (year) 1.045 (1.020e1.071) <0.001 GCS at admission 0.848 (0.735e0.978) 0.024 One or both non-reacting 2.040 (0.889e4.682) 0.093 pupils at admission Subarachnoid hemorrhage 3.487 (0.944e12.881) 0.061 Intraventricular hemorrhage 0.381 (0.080e1.802) 0.223 Epidural hemorrhage 0.340 (0.091e1.272) 0.109 Status of basal cisterns Normal 1.000 Compressed 4.643 (0.981e21.970) 0.053 Absent 6.136 (1.101e34.214) 0.039 Midline shift (mm) 1.092 (1.006e1.185) 0.036 Timing of cranioplasty (day) 1.005 (0.997e1.013) 0.195 Fig. 1. Distribution of timing of cranioplasty following decompressive craniectomy. ORIGINAL RESEARCH

Y.-H. Huang et al. / International Journal of Surgery 11 (2013) 886e890 889

The repair of skull defects is usually for protective and cosmetic need for additional hospital admissions. In this series, we examined purposes. From the pathophysiological point of view, cranioplasty the incidence of neurosurgical site infection and hydrocephalus appears to have a marked effect on postural blood flow regulation, based on the timing of cranioplasty. The results also showed that cerebrovascular reserve capacity, and cerebral glucose meta- the difference in the infection rate based on different time intervals bolism,18 but a review of the clinical literature reveals controversy was trivial. On the other hand, there seems to be a trend toward a with regard to the curative effects of cranioplasty. Improvements in higher rate of hydrocephalus in patients with late cranioplasty. general cognitive function and motor deficits following cranial Since our study was aimed at the neurological outcomes of TBI, reconstruction have been described in a few case reports.19,20 further suitably-designed studies would be required to avoid se- Furthermore, Stiver et al. examined 38 patients with long-term lection bias before drawing any solid conclusions about the influ- follow-up after DC for TBI. They found delayed monoparesis in 10 ence of the timing of cranioplasty on complications. patients and reversal of the weakness following cranioplasty repair.21 In contrast, Heidi et al. collected 23 patients undergoing 5. Conclusion cranioplasty for prior DC, including 15 trauma and 8 non-trauma cases.22 The authors presented no evidence that cranioplasty has Our results show that the timing of cranioplasty following an effect on GCS in addition to the expected time-based recovery. posttraumatic DC is not related to the neurological outcomes of TBI. There is doubt that the documented benefits of cranioplasty may Despite the limitations of the retrospective design, the analyses partially or completely result from spontaneous neuronal provide preliminary information to elucidate the question. improvement. Actually, severe TBI causes dynamic processes to commence or change over time from injury to recovery, and the Ethical approval importance of the timing of surgical intervention should be The study protocol was approved by the Institutional Review emphasized. Committee on Human Research at the Chang Gung Memorial The optimal timing from injury to DC is controversial, but is Hospital. accepted to be within 24 h after injury.17 The rationale for early decompressive surgery is based on improving cerebral perfusion Funding and preventing ischemic damage. As for cranioplasty surgery, the The authors declare no source of funding for this study. role of the time interval from DC to cranial repair in determining prognosis is not clear. To the best of our knowledge, this is one of Author contribution the few studies to focus on the relationship between the timing of Yu-Hua Huang: study design, writing. cranioplasty and neurological outcomes of TBI. Although the results Tao-Chen Lee: data analysis. show that no specific time frame was predictive, this data provides Ka-Yen Yang: data collections. additional information for clinicians to assist with therapy- Chen-Chieh Liao: data collections. planning. We consider that the timing of cranioplasty should not be viewed as a determinant of the neurological outcome. In most Conflict of interest studies, the timing is usually used as a cut-off value for early and The authors declare no potential conflicts of interest relevant to late cranioplasty.3,5,23 We have analyzed the association between this article. timing and outcome in a continuous way, and believe that a fairly relation across time frame approximates the reality. However, there exist limitations in the evaluation of neurological outcomes. The References GOS was used because of its widespread availability, but it may be insensitive to functionally significant differences in performance. 1. Polin RS, Shaffrey ME, Bogaev CA, et al. Decompressive bifrontal craniectomy in Moreover, the GOS is pseudo-ordinal and dichotomized into the treatment of severe refractory posttraumatic cerebral edema. Neurosurgery 1997;41:84e92 [discussion e4]. favorable and unfavorable outcomes, which further reduces the 2. Cooper DJ, Rosenfeld JV, Murray L, et al. Decompressive craniectomy in diffuse sensitivity. Other disadvantages of this study were related to its traumatic brain injury. N Engl J Med 2011;364:1493e502. fi retrospective design, as follows: The assessment of neurological 3. Chun HJ, Yi HJ. Ef cacy and safety of early cranioplasty, at least within 1 month. J Craniofac Surg 2011;22:203e7. status is less complete and accurate than in planned research. The 4. Beauchamp KM, Kashuk J, Moore EE, et al. Cranioplasty after postinjury period of evaluation of neurological outcomes varies and follow-up decompressive craniectomy: is timing of the essence? J Trauma 2010;69:270e4. information is difficult to obtain without scheduled returns to the 5. Cheng YK, Weng HH, Yang JT, Lee MH, Wang TC, Chang CN. Factors affecting graft infection after cranioplasty. J Clin Neurosci 2008;15:1115e9. clinic. A number of patients recover neurologically more than 6 6. Lingsma HF, Roozenbeek B, Steyerberg EW, Murray GD, Maas AI. Early prog- months after TBI, but long-term results may be absent in the nosis in traumatic brain injury: from prophecies to predictions. Lancet Neurol evaluation. 2010;9:543e54. 7. Sahuquillo J, Arikan F. Decompressive craniectomy for the treatment of re- The timing of cranioplasty has generally been discussed because fractory high intracranial pressure in traumatic brain injury. Cochrane Database of complications including infection or hydrocephalus in the Syst Rev 2006:CD003983. interim period.24,25 Investigators have historically believed that 8. Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG. CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of early cranioplasty is associated with greater morbidity resulting surgical wound infections. Infect Control Hosp Epidemiol 1992;13:606e8. from insufficient scalp conditions or unresolved trauma insult, but 9. Korinek AM. Risk factors for neurosurgical site infections after craniotomy: a recent publications have espoused an alternative view.23 Chun et al. prospective multicenter study of 2944 patients. The French Study Group of found that early cranioplasty (within 1 month after DC) provides a Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Service Epi- demiologie Hygiene et Prevention. Neurosurgery 1997;41:1073e9 [discussion satisfactory securing dissection plane during operative procedures, 9e81]. without causing additional complications, such as infection, sub- 10. Murray GD, Butcher I, McHugh GS, et al. Multivariable prognostic analysis in 3 traumatic brain injury: results from the IMPACT study. J Neurotrauma 2007;24: dural hygroma, and brain parenchymal damage. Beauchamp et al. e e 329 37. reported delayed cranioplasty (3 6 months post-DC) does not 11. Tokutomi T, Miyagi T, Ogawa T, et al. Age-associated increases in poor out- seem to lower post-cranioplasty infection rates or the need for comes after traumatic brain injury: a report from the Japan Neurotrauma Data cerebrospinal fluid diversion procedures.4 They emphasize per- Bank. J Neurotrauma 2008;25:1407e14. 12. Aarabi B, Hesdorffer DC, Ahn ES, Aresco C, Scalea TM, Eisenberg HM. Outcome forming cranioplasty as soon as there is resolution of brain swelling following decompressive craniectomy for malignant swelling due to severe on the CT scan, to lower the overall cost of care by eliminating the head injury. J Neurosurg 2006;104:469e79. ORIGINAL RESEARCH

890 Y.-H. Huang et al. / International Journal of Surgery 11 (2013) 886e890

13. Jiang JY, Xu W, Li WP, et al. Efficacy of standard trauma craniectomy for re- 19. Sancisi E, Battistini A, Di Stefano C, Simoncini L, Montagna P, Piperno R. Late fractory intracranial hypertension with severe traumatic brain injury: a recovery from post-traumatic vegetative state. Brain Inj 2009;23:163e6. multicenter, prospective, randomized controlled study. J Neurotrauma 20. Segal DH, Oppenheim JS, Murovic JA. Neurological recovery after cranioplasty. 2005;22:623e8. Neurosurgery 1994;34:729e31 [discussion 31]. 14. Ucar T, Akyuz M, Kazan S, Tuncer R. Role of decompressive surgery in the 21. Stiver SI, Wintermark M, Manley GT. Reversible monoparesis following management of severe head injuries: prognostic factors and patient selection. decompressive hemicraniectomy for traumatic brain injury. J Neurosurg J Neurotrauma 2005;22:1311e8. 2008;109:245e54. 15. Ho KM, Honeybul S, Litton E. Delayed neurological recovery after decom- 22. Stelling H, Graham L, Mitchell P. Does cranioplasty following decompressive pressive craniectomy for severe nonpenetrating traumatic brain injury. Crit craniectomy improve consciousness? Br J Neurosurg 2011;25:407e9. Care Med 2011;39:2495e500. 23. Yadla S, Campbell PG, Chitale R, Maltenfort MG, Jabbour P, Sharan AD. Effect of 16. Pompucci A, De Bonis P, Pettorini B, Petrella G, Di Chirico A, Anile C. Decom- early surgery, material, and method of flap preservation on cranioplasty in- pressive craniectomy for traumatic brain injury: patient age and outcome. fections: a systematic review. Neurosurgery 2011;68:1124e9 [discussion 30]. J Neurotrauma 2007;24:1182e8. 24. Sobani ZA, Shamim MS, Zafar SN, et al. Cranioplasty after decompressive cra- 17. Howard JL, Cipolle MD, Anderson M, et al. Outcome after decompressive craniectomy: an institutional audit and analysis of factors related to complica- niectomy for the treatment of severe traumatic brain injury. J Trauma 2008;65: tions. Surg Neurol Int 2011;2:123. 380e5 [discussion 5e6]. 25. Tahir MZ, Shamim MS, Sobani ZA, Zafar SN, Qadeer M, Bari ME. Safety of un- 18. Winkler PA, Stummer W, Linke R, Krishnan KG, Tatsch K. Influence of cranio- treated autologous cranioplasty after extracorporeal storage at 26 degree plasty on postural blood flow regulation, cerebrovascular reserve capacity, and celsius. Br J Neurosurg 2013 Jan 7. [Epub ahead of print] PubMed PMID: cerebral glucose metabolism. J Neurosurg 2000;93:53e61. 23293976.